Brain arteriovenous malformations (BAVM) are usually detected only when the patient presents with intracranial hemorrhages, seizures, headaches, focal neurological deficits, or other disorders. A smaller number are detected incidentally. Although BAVM themselves are rare, the severity of the clinical presentations underscores the need for a better understanding of etiology, natural history, and prognosis of the disorder. The primary reason for treating BAVM is the prevention of intracranial hemorrhage. However, the risk associated with treatment must be weighed against that of natural history of BAVM, on which there is little data. In addition, there is controversy involving the relative efficacies of the 2 most practiced treatment modalities: open surgery and stereotactic radiosurgery. Most of what is known about BAVM comes from referral series. There is a paucity of data from population-based studies. In this project, we will examine patient and BAVM characteristics in relation to BAVM presentation and subsequent hemorrhage using both referral-based and population-based data. We will especially focus on the epidemiological issues involved in BAVM research methodologies. Our project will also investigate the settings for clinical equipoise in BAVM treatment using an improved observational study design. Results of this latter analysis will significantly contribute to the planning of a future randomized clinical trial for BAVM treatment, comparing open surgery with radiosurgery. This proposed project will serve as a critical training experience for the Principal Investigator, by affording him the opportunity to apply his epidemiological background and augment his research skills in preparation for independent scientific research. The results of this project may additionally further our understanding of the natural history and risk factors for intracranial hemorrhages in patients with BAVM, as well as evaluate the safety and efficacy of the different treatment modalities.